On Sat, 4 Sep 2004, Laurent Pinchart wrote:

> > Maybe we're talking about different traces.  I'm referring to the trace
> > for the Zoom modem (a low-speed device according to Steve), as attached to
> 
> Yes we are. The Zoom modem is a USB 1.0 device according to Steve, but he 
> didn't mention if it is a low-speed or high-speed device (well if he did I 
> didn't notice).
> 
> Steve sent me another trace which he didn't post to the list due to its size. 
> You can find the zipped file at http://ben.lautard.free.fr/mouse_enum.txt.gz 
> (I asked a friend to host it for me, so please download it as soon as you 
> can, it will probably not stay there for very long).

Okay, I got it.  Thanks.

> The trace shows the enumeration of a USB 1.1 low-speed HID device. It shows 
> (not very clearly, as its very verbose, but still) that Windows XP enumerates 
> the low-speed device with a 18 bytes Get-Device-Descriptor request.

Hmm...  It's a little tricky to decipher this, as you say.  At the start
is a bunch of Events labelled "Keep_Alive".  Just after Event 32 there's
something labelled "Transfer Header" -- I suppose it is the decode of the
message that follows.  Next I see a Transaction Header for Transaction
Number 0; this must be the start of the first Get-Device-Descriptor
control message.  It shows a Setup packet, then a DATA0 packet, then an
ACK.  In the DATA0 packet the 7th byte is 0x40; hence the request is for
64 bytes.

The device transfers all 18 bytes in three transactions, appropriate for a 
low-speed device with maxpacket = 8.  Finally, Event 35 is a bus reset.

Alan Stern



-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to